This invention relates to the preparation of therapeutic compositions consisting of purified blood coagulation factors for the treatment of haemostatic disorders. Said compositions are obtained by affinity chromatography employing antibodies, more in particular monoclonal antibodies, that distinguish between intact and cleaved molecular species. Methods are disclosed to obtain such antibodies, which allow the isolation of vitamin K-dependent blood coagulation proteins, including Factor IX, Factor VII, Protein C or Protein S, as intact polypeptides, devoid of cleavage products representing activated or degraded species.
Inherited or acquired deficiencies of proteins of the blood coagulation system provide a major cause for the occurrence of haemostatic disorders. Even the lack or shortage of one single component of this system may be sufficient to disturb the delicate balance between procoagulant and anticoagulant pathways in a manner resulting in major clinical signs of bleeding or thrombosis. One of the most common bleeding disorders is Haemophilia A, which is due to deficiency or dysfunction of the coagulation Factor VIII. Less frequently occurring, but equally severe bleeding disorders include deficiencies of the haemostatic proteins Factor IX (Haemophilia B), Factor VII, or Factor X. On the other hand, thrombosis may occur as the result of even partial (heterozygous or acquired) deficiency of Protein C or Protein S, which are major components of a system that acts as an antagonist of the coagulation pathway (for reviews on haemostatic disorders see A. L. Bloom and D. P. Thomas (Eds.), Haemostasis and Thrombosis, 2nd edition, Churchill-Livingstone, Edinburgh, 1987, pp 393-436 and 452-464). Replacement therapy is considered as a powerful and effective means to restore the haemostatic balance in vivo. For instance, concentrates containing Factor IX have proven highly valuable blood products which are life-saving when used to control bleeding in patients suffering from Factor IX deficiency.
Commercially available Factor IX concentrates (so-called prothrombin complex concentrates) usually are prepared with ion exchange resins to separate Factor IX from the other plasma proteins. This technique however, yields Factor IX preparations that also contain a number of other, closely related haemostatic proteins. These include Factor VII, Factor X, Factor II, Protein C and Protein S, which all belong to the class of the vitamin K-dependent proteins. The term xe2x80x9cvitamin K-dependentxe2x80x9d is referring to the fact that these proteins contain glutamic acid residues that are carboxylated during biosynthesis in a vitamin K-dependent process. Carboxylation provides these proteins with unique Ca2+-binding sites that are obligatory for the biological activity of these proteins within the Ca2+-dependent haemostatic process. Due to these and other structural similarities (see B. Furie and B. C. Furie, Cell vol 53, 1988, pp 505-518), the vitamin K-dependent proteins are readily co-purified. Thus, most Factor IX concentrates are also containing other haemostatic proteins such as Factor VII, Protein C and Protein S, and consequently the same concentrates have been used for the treatment of deficiencies of those proteins as well. However, treatment of bleeding disorders with compositions containing anticoagulant proteins such as Protein C and Protein S is in fact highly undesired, as is treatment of thromboembolic disease with compositions containing Factor IX, Factor VII or other procoagulant components as major contaminants. Therefore, the ideal therapeutic composition to correct a deficiency of one specific haemostatic protein should consist of solely that single component in an intact conformation and nothing else except solvent, and sometimes an inert carrier. As a consequence, the purification strategies needed to achieve the desired degree of purity have become increasingly complex.
Along with the introduction of advanced, more complex purification protocols a novel problem of undesired proteolysis of the target species was encountered. Limited proteolysis is a key mechanism in the regulation of a number of biological systems (see H. Neurath and K. A. Walsh, Proc. Natl. Acad. Sci. USA vol 73, 1976, pp 3825-3832). Typical examples of such biological systems include the complement system, the fibrinolytic system, and the blood coagulation system. These biological cascade systems involve the sequential conversion of intact, inactive precursor proteins into active enzymes or cofactors by proteolysis of one or more specific peptide bonds. On the other hand, feedback mechanisms exist to maintain these processes under local control and lead to proteolytic inactivation of the target proteins. Accordingly, the components of the coagulation cascade are present in blood plasma in a precursor form lacking biological activity. With regard to replacement therapy, the presence of coagulation proteins that are no longer intact is troublesome, since after having been subject to limited proteolysis, such cleaved species may bypass the natural, local control of haemostasis in vivo. Although natural mechanisms effectively control proteolysis under physiological conditions, these can no longer be maintained when haemostatic proteins are isolated from their natural source. As such protease-sensitive sequences are exposed within the tertiairy structure of these proteins, they provide easily accessible targets for proteolysis in a non-physiological environment lacking natural control mechanisms. Therefore, it is virtually impossible to completely prevent partial proteolysis during purification. Uncontrolled proteolysis of these vulnerable proteins is not limited to purification from a natural source as human plasma or fractions thereof, but may equally occur when the same proteins are obtained by recombinant DNA technology from transformed cell lines in vitro, or from biological fluids, including milk, of transgenic animals in vivo.
The presence of cleaved species in therapeutic products is clearly not desired, because the presence of activated proteins may trigger thrombogenic responses of the haemostatic system, whereas the presence of inactivated proteins leads to products with suboptimal biological activity that may competitively inhibit the reactions to be corrected. Prothrombin complex concentrates contain activated species of virtually all vitamin K-dependent coagulation factors, and this has been established as a causative agent for the occurrence of thromboembolic complications since the 1970s is (G. C. White et al., Blood vol 49, 1977, pp 159-170; J. M. Lusher, Seminars in Hematology vol 28, suppl 6, 1991, pp 3-5). Theoretically, in particular those species that participate in the initiation phase of the coagulation system, and thereby are the most amplified in the cascade mechanism, are to be considered as the most potent in disturbing the physiological haemostatic balance. Indeed, in vivo studies employing purified activated coagulation factors have identified activated Factor IX (S. Gitel et al., Proc. Natl. Acad. Sci. USA vol 74, 1977, pp 3028-3032) and activated Factor VII (K. Mertens et al., Thromb. Haemostasis vol 64, 1990, pp 138-144) as thrombogenic even in extremely low dosage. This may raise particular concern for activated forms that are relatively resistent to inhibition in vivo. Most activated vitamin K-dependent coagulation factors are subject to almost instantaneous inhibition by the abundance of protease inhibitors in blood plasma. However, Factor IXa is only slowly inhibited, whereas Factor VIIa and activated Protein C have in vivo half-lives up to 2 hours (K. Mertens et al., Thromb. Haemostasis vol 64, 1990, pp 138-144; P. C. Comp, Hematology, McGraw-Hill, New York, N.Y., 1990, pp 1290-1303). Thus, upon infusing vitamin K-dependent haemostatic proteins into patients, it should be noted that even traces of activated forms may remain in the patients circulation sufficiently long to bypass physiological control. Therefore, Protein C and Factors VII and IX are among the proteins that should be prevented from being activated, or should be purified employing a strategy that is selective for their intact zymogens. On the other hand, other cleavages can occur that result in the inactivation of said proteins, thus reducing their therapeutic efficacy. For instance, Factor IX can be inactivated by the enzymes thrombin or elastase, and the cleavage product has no longer the potential of being converted into a form having Factor IXa activity (A. Takaki et al., J. Clin. Invest. vol 72, 1983, pp 1706-1715; W. Kisiel et al., Blood vol 66, 1985, pp 1302-1308). Similarly, the anticoagulant Protein S is readily cleaved by thrombin into a product that no longer has anticoagulant activity (for review see M. Hessing, Biochem. J. vol 277, 1991, pp 581-592). It thus appears highly important to avoid the occurrence of cleaved, non-intact species to reduce side-effects and to improve efficacy of therapeutic concentrates of these factors.
To better appreciate the structural differences between the intact zymogen species and their cleaved derivatives, the molecular events associated with limited proteolysis of these proteins are now described in more detail. Table I presents an overview of target sequences for limited proteolysis as they occur in a number of vitamin K-dependent proteins involved in the haemostatic system. These include:
(a) Human Factor VII: The intact zymogen is a single-chain glycoprotein of 406 amino acids, which is converted to Factor VIIa by the cleavage of a single bond between residues 152 and 153 (see Table I). This target sequence can be cleaved by a number of enzymes, including thrombin and Factors IXa, Xa and XIIa, which results in the activated species which consists of two polypeptide chains that are held together by a disulfide bond. This Factor VIIa provides a powerful thrombogenic trigger by activating a number of target substrates, including Factor X and Factor IX (B. Osterud and S. I. Rapaport, Proc. Natl. Acad. Sci. USA vol 74; 1977, pp 5260-5264).
(b) Human Factor IX: The intact zymogen is a single-chain glycoprotein of 415 amino acids, which is converted into Factor IXa by Factor XIa or Factor VIIa in two steps. The first step involves the cleavage between residues 145 and 146, which results in the formation of a two-chain inactive intermediate (called Factor IXxcex1) containing a light chain of residues 1-145 and a heavy chain of residues 146-415. In the second step, the heavy chain of Factor IXxcex1 is further cleaved between residues 180 and 181, resulting in a 35-residue (146-180) activation peptide and an active enzyme (Factor IXaxcex2) with the remaining heavy chain of residues 181-415. It should be noted that whereas the latter cleavage is required to develop the final Factor IXa activity, the prior cleavage of the 145-146 bond seems to be an obligatory step in Factor IX activation by Factor XIa or Factor VIIa (U. Hedner and E. W. Davie, in: R. W. Colman et al. (Eds), Hemostasis and Thrombosis, J. B. Lippincott, Philadelphia, 1987, pp 29-38; M. J. Griffith et al., J. Clin. Invest. vol 75, 1985, pp 4-10). Therefore, Factor IXxcex1 is even more vulnerable than the intact Factor IX zymogen with respect to cleavage at the 180-181 position and the concurrent formation of a thrombogenic Factor IXa species. In addition to this proteolytic activation, also inactivation may occur as the result of cleavage by thrombin or elastase. This proteolytic inactivation involves cleavage between residues 327 and 328, and between residues 338 and 339, resulting in derivatives that can no longer be converted to species with Factor IX procoagulant activity (see Table I).
(c) Human Protein C: The intact zymogen is a two-chain glycoprotein of 419 amino acids, consisting of a light chain of residues 1-155 and a heavy chain of residues 158-419. The zymogen is activated by a single cleavage in the heavy chain, between residues 169 and 170. The activated species is a powerful anticoagulant, which inactivates Factors Va and VIIIa in a manner requiring the presence of a vitamin K-dependent cofactor, called Protein S.
(d) Human Protein S: The intact species is a single-chain glycoprotein of 635 amino acid residues. The N-terminal portion of the protein contains two thrombin-sensitive bonds between residues 49-50 and 70-71. After cleavage of Protein S by thrombin, the N-terminal fragment still is connected to the molecule via a disulfide bond. However, only Protein S that is uncleaved within the thrombin-sensitive region is active as a cofactor for activated Protein C. Therefore, it is preferred to provide intact, uncleaved Protein S in a composition for effective treatment of heriditary or acquired Protein S deficiency.
In conclusion, a special need exists for purification methods that allow the selection of uncleaved, intact species from a source containing the intact proteins as well as proteolytic derivatives generated by cleavage at the target peptide bonds that are summarized in Table I.
The development of monoclonal antibody technology has radically changed the potential for purifying trace proteins from plasma to near homogeneity. Monoclonal antibodies can be prepared and selected in such a manner that they show complete specificity towards an individual plasma protein. When a monoclonal antibody specific for a plasma protein is immobilized on an inert matrix, it will absorb that plasma protein specifically from a mixture containing that plasma protein. After washing the matrix, the plasma protein may be eluted under fairly mild conditions in pure or virtually pure form. This approach has allowed the isolation of pure Factor IX from human blood plasma by immuno-affinity chromatography (A. H. Goodall et al., Protides of the Biological Fluids (Peeters, Ed.) vol 30; Pergamon Press, Oxford, 1982, pp 403-407; K. Mertens et al., Thromb. Haemostasis vol 50; 1983, p 249; K. Mertens, Ph.D. dissertation, State University of Leiden, 1985, pp 83-98). Similar methods of immuno-affinity chromatography have subsequently become available for purification of the other vitamin K-dependent haemostatic proteins.
In some cases, a more restricted specificity has been achieved, in that antibodies may specifically bind to their target vitamin K-dependent protein in the presence of Ca2+-ions. Under these conditions, the target proteins are in a biologically active conformation by virtue of the interaction of Ca2+-ions with the carboxylated glutamic acid residues that are unique to the vitamin K-dependent proteins. When applied in a purification process, such antibodies allow elution of the target protein under mild conditions by the Ca2+-binding agent EDTA, as has been described for Factor IX (H. A. Liebman et al., Proc. Natl. Acad. Sci. USA vol 82, 1985, pp 3879-3883; K. J. Smith, Blood vol 72, 1988, pp 1269-1277) as well as for Protein C (D. J. Stearns et al., J. Biol. Chem. vol 263, 1988, pp 826-832).
However, in spite of their high selectivity, those procedures have not been able to distinguish between the intact and activated or degraded coagulation factors, and as a consequence cleaved products are co-purified with the desired intact, non-activated target product. Moreover, antibodies that are Ca2+-dependent have the disadvantage that Ca2+-ions cannot be added to most source materials, including plasma or fractions thereof, without simultaneously triggering the Ca2+-dependent coagulation system that leads to proteolytic cleavage of the vitamin K-dependent target proteins. This represents a major drawback for the applicability of the final therapeutic product as it has been demonstrated that even small amounts of activated coagulation factors have been implicated as causative agents for disseminated intravascular coagulation and thromboembolism.
The present invention relates to methods for the generation and selection of Ca2+-independent, monospecific antibodies (either polyclonal or monoclonal) specific for defined epitopes covering intact (activation or degradation) cleavage sites of haemostatic proteins such as, for example, coagulation Factor IX and Protein S, for the isolation of these proteins as intact, uncleaved polypeptides. The strategy of this method is based on the notion that disruption of the primary sequence of the target protein leads to a number of changes within the protein, one being the exposition of the cleavage site itself. The present invention demonstrates that antibodies against epitopes covering intact cleavage sites may have different affinity for intact and cleaved forms of the epitope, thus allowing separation when applied in a chromatographic process. In order to be applicable in purification processes from a variety of sources, the antibodies described in the present invention are not only specific for a sequence that provides a target for limited proteolysis within the coagulation factor to be isolated, but are also Ca2+-independent. The method involves the selection of antibodies with appropriate specificity, immobilization thereof on a solid support, and then contacting a source material with the immobilized antibody to bind said protein. After washing and elution with an appropriate solution, the target protein is recovered in a highly purified form, in a non-activated, intact state that is of particular interest for use in a therapeutic composition for the treatment of haemostatic disorders.
The process of the present invention is a major breakthrough in the field of immuno-affinity chromatography in that it achieves unique selectivity for intact, non-cleaved target proteins. This allows for methods to obtain intact haemostatic proteins, which can thereafter be formulated into improved therapeutic blood products. These products provide safer and more effective agents than are available thus far for the treatment of patients encountering critical bleeding or clotting episodes.
Other objects and advantages of the present invention will become evident from the following detailed description of the invention.